The Discovery Sprint That Pays for Itself: Pricing Pre-Build Work Without Losing the Deal

Every agency has done it: three weeks of unpaid "scoping calls," a 40-page proposal, and then the prospect ghosts. Meanwhile the sales pipeline looks healthy on the dashboard and the P&L quietly bleeds. The fix isn't better proposals — it's charging money before you write one.

A paid discovery sprint reframes the sale. Instead of guessing at scope for free and hoping to win, you sell a small, fixed-price engagement that produces a real artefact. Done right, it closes more build deals, filters tyre-kickers, and — this is the part nobody tells you — is profitable on its own.

Why free scoping is a losing game

The economics of unpaid discovery are brutal. If your solutions engineer spends 30 – 40 hours on a proposal and you close one in four, your fully-loaded cost of sale on a lost deal can approach five figures. Multiply by ten prospects a quarter and you're funding a small engineering team just to write documents nobody reads.

Worse, free scoping trains the client to treat your expertise as a commodity. If the first thing you give away is architectural thinking, don't be surprised when they push back on the invoice for the same thinking three months later.

There's also a quality problem. A free scope is a rushed scope. You skip user interviews, skip the technical spike on the third-party API that will absolutely break in production, and skip the honest conversation about what "real-time" actually means. Then you sign a fixed-price contract on top of a fantasy — and eat the difference.

What a discovery sprint actually is

A discovery sprint is a paid, time-boxed engagement — usually one to three weeks — where you produce the artefacts needed to quote and start a build with confidence. It is not a workshop. It is not a strategy deck. It is engineering-grade due diligence with deliverables the client owns whether or not they hire you for the build.

Typical deliverables:

• A prioritised feature list mapped to user outcomes
• A technical architecture diagram with named services and third-party dependencies
• A risk register (integrations, data migration, compliance, unknowns)
• A milestone-based build plan with ranges, not point estimates
• A fixed-price or T&M proposal for the build phase

The "walk-away" clause

This is the trust move. Write into the SOW that the client owns everything produced during discovery and is under no obligation to proceed with you for the build. Say it out loud on the sales call. It disarms the "you're just trying to lock us in" objection and it forces you to make the artefacts genuinely portable.

Counter-intuitively, offering the walk-away increases build conversion. Prospects who feel free to leave rarely do, because by week two your team has out-thought their internal team and the switching cost is obvious.

Pricing the sprint

The number needs to be big enough to be taken seriously and small enough to be a rounding error against the build. Our rule of thumb: 5 – 10% of the estimated build budget, with a floor.

A rough ladder we've used:

• Small (£8k – £15k, one week): single product surface, one or two integrations, a clear "we know what we want" client
• Standard (£20k – £35k, two weeks): a new product from scratch, or a rebuild with unclear requirements
• Enterprise (£40k – £75k+, three to four weeks): multi-system integration, compliance considerations, procurement involved

Don't discount. If the prospect balks at the sprint fee, they will balk at every change request in the build. That's data, not rejection.

What to bill against

Fixed price. Always. The sprint's whole purpose is to demonstrate you can scope and deliver a small commitment on time — if you can't do that, why should the client trust you with a six-month build? Bill 50% on signature and 50% on delivery of the artefacts.

Staffing it properly

The biggest mistake agencies make is running discovery with junior BAs and a PM. A discovery sprint needs your best people, in short bursts:

• A senior engineer (30 – 50% allocation) to spike risky integrations and write the architecture doc
• A product lead (50 – 70%) to run interviews and shape the backlog
• A designer (20 – 30%) for enough wireframes to make the scope concrete
• A delivery lead (10 – 20%) to build the milestone plan and quote the build

Yes, this is expensive. That's why the fee isn't small. But the same team that runs discovery should lead the build — continuity is a huge part of what the client is buying.

The week-by-week shape

Here's a two-week sprint plan we've used repeatedly. Adapt the cadence, keep the artefacts.

Week 1
  Day 1  Kick-off, stakeholder map, success criteria
  Day 2  User interviews (3–5), competitive teardown
  Day 3  Feature workshop, MoSCoW prioritisation
  Day 4  Technical spike on riskiest integration
  Day 5  Draft architecture, draft backlog

Week 2
  Day 6  Wireframes for critical flows
  Day 7  Risk register, non-functional requirements
  Day 8  Milestone plan, effort ranges
  Day 9  Internal review, pricing the build
  Day 10 Deliverable walkthrough with client

The Day 10 walkthrough is where you present the build proposal. Not by email. In the room, with the artefacts open, and the number on screen. Close rates jump when the proposal is the natural conclusion of a shared workshop rather than a PDF that arrives cold on a Tuesday afternoon.

Contracts and IP

Keep the SOW under five pages. State the fee, the timeline, the named deliverables, the walk-away clause, and IP assignment. Everything the client is paying for is theirs on receipt of final payment. Your background IP (frameworks, templates, internal tools) stays yours.

One clause worth adding: a build-phase discount if they proceed within, say, 60 days. We've offered 50% of the sprint fee credited against the first build invoice. It creates urgency without discounting the build itself, and it makes the sprint feel like a deposit rather than a tax.

Handling the "can't you just quote it?" objection

You will hear this on every third call. The honest answer is: yes, we can, and it will be wrong by a factor of two in either direction, and one of us will be unhappy about it. Say that. Then offer the sprint as the alternative to the guess.

A line that works:

We can give you a rough range today — somewhere between £X and £3X. If you need it tighter than that, we need two weeks and a proper look. The sprint is how we get you a number you can actually put in a board pack.

Prospects who are serious about building something respect this. Prospects who wanted a free architecture review self-select out, which is exactly the outcome you want.

Measuring whether it's working

Track three numbers:

1. Sprint sell-through rate: qualified opportunities that buy a sprint. Aim for above 40% within six months of introducing this.
2. Sprint-to-build conversion: sprints that convert into a build within 90 days. Above 70% means your sprint is doing its job. Below 50% means you're either scoping the wrong problems or your build pricing is out of market.
3. Build margin on sprint-originated projects: should be materially higher than cold-quoted work. If it isn't, your discovery process is theatre.

If conversion is high but margin isn't improving, the sprint isn't producing honest estimates — it's producing optimistic ones because the team wants the follow-on work. Fix the incentive by having a separate delivery lead price the build.

Where we'd start

Pick your next three inbound leads. Quote a discovery sprint instead of writing a proposal. Use the fee ladder above, insist on fixed price, include the walk-away clause. If all three say no, your pipeline was weaker than you thought — better to know now. If one says yes, you've just replaced 120 hours of unpaid proposal work with a paid, profitable engagement that ends with a build contract on the table. That's the whole game.

If you want a second pair of eyes on how you're currently scoping and pricing, our product strategy work starts exactly here.